I.
PHYSICAL ELECTRONICS
Prof. W. B. Nottingham
H. S. Jarrett
J. M. Bailey
D. Jeffries
A. R. Hutson
L. E. Sprague
R. T. Watson
A.
ELECTRON-EMISSION PROBLEMS
1.
Thermionic Work Function and Conductivity of Oxide-Coated Cathodes
The investigation of the thermionic emission and conductivity of an oxide-coated
cathode is being continued in order to determine the true values of the energy levels in
(Ba, Sr)O.
With the cathode activated on the vacuum system it yielded a thermionic work
function of 1.38 ev and a conductivity activation energy of 1. 31 ev.
was made to increase the activation of the cathode.
was removed from the vacuum system.
Another attempt
When this was completed, the tube
With the Bayard-Alpert ion gauge running, the
straight-line portion of a Richardson plot resulted in
4
= 0.98 ev.
In the low-temperature region of Richardson plots taken at this state of activation,
the observed electron emission decreased with time.
suspected.
Contamination of some sort was
To measure the contamination involved, observations were made of the
decrease in emission of one of the ion
gauge filaments run at 1660"K as a function
of the time after flashing the filament to
2200'K.
Contamination was evident even
though the pressure remained constant at
1.5 x 10
-8
-
from day to day.
The getter
was flashed at this time with the ion gauge
inoperative.
One-half hour after turning
on the ion gauge the rate of contamination
ýT2
LOG
had decreased.
Furthermore, after
15 hours of gauge operation, there was a
very marked decrease in the contamination
rate.
-15
The conclusion is that both an oper-
ating ion gauge and a flashed getter were
needed to clean up the contaminating gases.
A probable explanation of this cleanup
action is that the gauge breaks up molecules or ionizes molecules and atoms
1.5
20
which then react readily with the getter.
An experiment was done with the ion
25
gauge which showed that positive ions were
Fig. I-1
Richardson plot.
flashed off a hot tungsten filament which
-1-
(I.
PHYSICAL ELECTRONICS)
had been resting at room temperature for 13 days. The positive ion current in this case
was at least 1. 5 x 10 - 7 amp. This is interesting because only a few atoms and molecules have ionization potentials close to the work function of tungsten. This result has
been found and is recorded in the work on ionization gauges elsewhere in this report.
After the getter was flashed, research on Richardson plots in the temperature range
from 7000K to 295 0 K revealed several interesting results. The low-temperature
deviation observed before the getter flash was no longer present. Temperature cycling
and jumping from low to high to low temperatures as shown in Fig. I-i was carried out
in order to show that straight lines can be obtained and yet be unreliable. Whether or
not the ion gauge was operating made a definite difference. This indicated that the emission from the oxide depended both on the particular gases present in the tube and
-7
-9
mm
to 10
whether or not the gases were ionized or disassociated at pressures of 10
Hg.
Research is being done on the photoelectric effect caused by the ion gauge filament.
R. T. Watson
The results of this work are not yet conclusive.
2.
Magnetic Velocity Analyzer Investigation of Thermionic Emission
from Tungsten
The contemplated measurements of the thermionic work function and its temperature
derivative are to be made on a recrystallized tungsten filament. It is required that the
filament have a single crystal occupying the entire cross section of the wire and
extending at least 2 cm in length. One such filament has been obtained, and the techniques have been mastered thoroughly enough so that these filaments may be prepared
as needed.
The schedule of preparation for such a filament is as follows:
A length of 0.003-inch diameter G.E. No. 218 "Non-sag" is ground and polished
until it appears as a smooth cylinder microscopically. The final diameter is 0. 002 inch.
2. The wire is then spring-mounted between press leads at the ends of a cylindrical
1.
glass envelope.
The envelope is coated inside with a ZnS phosphor, and is provided
with a cylindrically symmetrical wire structure which serves to collect the secondary
electrons from the phosphor in step No. 4.
3.
The wire is then heated by conduction.
Starting at 1700"K its temperature is
raised 100" at one-hour intervals until it has reached 2500 0 K where it is given one-half
hour.
4.
A potential of 5000 volts is then applied to the collector, and with a filament
temperature of about 2300 °K, the thermionic emission pattern characteristic of the
single crystals that have grown in the filament is observed. This pattern may be photographed with good contrast by using a Wratten C-4 deep blue filter.
A. R. Hutson
-2-
(I.
3.
PHYSICAL ELECTRONICS)
Photoelectric Emission
Some preliminary retarding potential and spectral energy distribution data have been
obtained on the molybdenum surfaces to be used as the substratum for the deposition of
germanium films in the experimental tube described previously. In order to obtain these
data, ultraviolet spectral lines from cadmium, mercury, and tellurium discharge tubes
described in the last report were used. The ultraviolet lines generated are as follows:
2001A, 2064A, 2142A, 2208A, 2259A, 2288A, 2385A, 2378A, 2399A, 2482A, 2537A,
These lines are sufficiently close
2345A, 2654A, 2698A, 2760A, 2804A and 2893A.
together to permit a thorough investigation of a wide spectral range. All but the 2001A
and 2064A lines were used in the data.
The molybdenum surfaces were
A
b?
ron-I
Uom
%JillCLLLl
lm
h
9
2000"K prior to the time the data were
taken. If the emission from a given
z8
U)
t
z
+1
+d+
to
spectral line is tabulated as a function
of time after heating, a curve similar to
that shown in Fig. I-2 for the mercury
2537A line is obtained. Apparently the
0_j6
w 2
zw
cr
cr
=
o
z0
l
1t
-4e
eL-aeC
y ec:
lu
5
decreasing part of the curve is the cooling
curve for the molybdenum cathode,
while the increasing portion is a gas
contamination curve.
3
The surface
seems to saturate after eight hours, and
the work function remains constant for
a period of days.
The pressure in the
2
0
2
3
4
5
6
7
8
9
10
tube as read by a Bayard gauge
HOURS AFTER BOMBING
6 x i0
Fig. 1-2 Rate of surface contamination.
is
mm Hg
mm Hg.
Figure I-3 shows the retarding poten-
tial curves obtained for the wavelength 2385A for a contaminated molybdenum surface
after it has reached the saturation value described above. The dashed curve represents
the duBridge (1) theory. Although not plotted, the theory of Mitchell (2), which takes
into account the transition probability of the electron, will lie between the experimental
and the duBridge curve.
To obtain a spectral energy distribution curve, or Fowler plot, the relative intensity
of the spectral lines must be known. These intensities were obtained from a radiation
thermocouple and a ZnSiO 4 phosphor painted on a 929 phototube. For the brighter lines
the thermocouple was used; for the dim lines the phosphor was used in conjunction with
the excitation spectrum found by Kr6ger (3). Figure I-4 shows the Fowler plot for the
The ordinate is in relative units proportional
molybdenum with saturated contamination.
-3-
--
3.2
r
~
rC
24
2.0
I I
I
I
I
I
-0.2
-0.1
0
0.80
040
-0.9
-0.8
-0.7
-0.6
-0.5
-0.4
-0.3
0.1
0.2
APPLIED ANODE POTENTIAL- VOLTS
Fig. I-3 Retarding potential plot X = 2385 A.
v,_j
W
8
hv IN ELECTRON VOLTS
Fig. I-4 Fowler plot for molybdenum work function 4.33 ev.
0.3
(I.
PHYSICAL ELECTRONICS)
to the number of electrons emitted per incident quantum. The cross represents the
work function of the molybdenum as 4. 33 ev, which is in close agreement with results
found by other experimenters (4).
References
Phys. Rev. 43, 727 (1933).
Proc. Camb. Phil. Soc. 31, 46 (1935).
1.
L. A. duBridge:
2.
K. Mitchell:
3.
F. A. Kriger:
4.
L. A. duBridge, W. W. Roehr:
Physica 6, 764 (1939).
Phys. Rev. 42, 52 (1932).
H. S. Jarrett
4.
Photoelectric Emission from Quartz
Since the last report the experimental tube for measuring the photoelectric properties of surface electrons in quartz has been completed. From the data taken several
conclusions can be drawn.
In the first place, the emission does occur and it can be measured by an apparatus
of this sort. As might be expected from the transparency of quartz in this region, the
photoelectric yield of the emission process is low. The currents induced in the electro-13
-14
meter circuit through the coupling capacitor were of the order 10
to 10
amp.
The threshold was not determined with the originally hoped-for precision, but apparently lies in the region 280 to 300 mp.
The corresponding work function is 4. 33 to 4. 13
ev.
In addition to the above photoelectric properties, some information was obtained
relating to the first cross-over potential of crystalline quartz. It was found that primary
electrons arriving with an energy of approximately 33 ev produced secondary electrons in a ratio greater than unity.
There was evidence that the actual cross-over
potential was not much less than this figure, though a limitation of this sort of experiD. Jeffries
ment is that, at best, it can only set an upper limit.
B.
STUDIES WITH GASEOUS DISCHARGE
1. Investigation of Low-Pressure Mercury Arcs
A program has been prepared for a continuation of the study of low-pressure mercury
J. M. Bailey
arcs. The work has not been undertaken.
C.
EXPERIMENTAL TECHNIQUES
1.
Spectral Emissivity of Tungsten
The latest tube constructed for this investigation has been completed except for its
evacuation. Although the evacuation has been started, it has been held up pending the
development of better vacuum gauges and other vacuum studies.
-5-
(I.
PHYSICAL ELECTRONICS)
2.
Ionization Gauge Studies
The vacuum gauge most commonly used for high vacuum measurement is the ionization gauge constructed with a hairpin filament along the axis of the tube and generally
used as the source of electrons.
Surrounding this filament and wound in the form of an
open helix is a second filament which serves as the electron collector.
The ion collector
is generally an open-ended cylinder more or less coaxial with the other two elements
and operated at a potential negative with respect to the electron-emission filament.
Although some photoelectric emission occurs at this ion collector as the result of the
light emitted from the hot filament this background current is small and easy to correct.
The "reverse photoelectric effect" or the X-ray effect sets a definite limit to the minimum gas pressure that can be measured for this gauge design.
This reverse photoeffect
was described in some detail in the Final Report,
1946,
June 30,
page 16.
The
phenomenon involves the emission of a photon from the electron collector when it
receives an electron from the filament. This photon in turn may be absorbed by the ion
collector and cause it to emit an electron which finally goes to the electron collector and
gives the same indication as the arrival of a positive ion at the ion collector.
This effect
is hard to distinguish from the arrival of positive ions although by making a detailed
stuay of the voltage-current cnaracteristics
of the gauge one can establish that the
observed current in the limiting case is the
result of reverse photoeffect only.
The first
major step in the elimination of this effect
was made by Alpert and Bayard (see Rev.
Sci. Instr. 21, 571, 1950).
Their method of
reducing the reverse photoeffect is to build
a gauge with an ion collector of very much
smaller cross section by placing it inside of
the electron collector.
The source of the
electrons is the filament outside the cylindrical electron collector.
The Bayard method
of solving the problem is definitely limited
to an imnrnvempnt of the order of one hundred
Fig. I-5
Ionization gauge.
Fl is electron
emitter; F 2 is electron collector;
S is shield (used as electron collector or at ion collector poten-
fold in the limiting range.
Since the ordinary
gauge is of very little use for a vacuum better
than 2 x 10 -8 mm Hg, the Bayard gauge is
-10
limited to 10
mm and higher pressures.
An entirely new gauge design has been
developed in connection with these studies
ial); C is ion collector.
-6-
(I.
PHYSICAL ELECTRONICS)
and is illustrated in Fig. 1-5. Two filaments F l and F 2 are shown mounted on a four-lead
press. One of these filaments serves as the electron emitter while the other one is the
electron receiver. These filaments are located at the center of a spherical bulb. The
ion collector is a small cylinder C and the disk S serves as a shield so that the light from
filament Fl cannot reach the collector C and the photons produced by electron absorption
at F 2 cannot go directly to the ion collector. The purpose of the spherical bulb is to
reflect all of this light back to the filament region. It should be noted that this gauge is
of an extremely simplified design and therefore much easier to build than either the commonly used gauge or the Bayard gauge. Two models of this gauge have been produced
When properly designed and operated this gauge has no
obvious limitation as regards the minimum pressure at which it can be used. The indications are that if the surface reflections from the glass wall are reduced sufficiently
Experiments reported in the October Quarterly Progress Report by Dr. M. K. Wilkenson
and others are well underway.
indicate that 10 - 12 mm can be obtained in suitably processed sealed-off tubes. The
experiments carried on so far with the new gauge shortly after it was sealed-off of the
-10
mm.
vacuum system indicate a pressure of 10
the vacuum system indicate a pressure of 10
mm.
The first gauge made according to this new design and operated as described above
showed a sensitivity equal to that of a good Bayard gauge. Two others were much lower
in sensitivity except when operated with the shield S functioning as an electron collector
as well as a shield.
Progress to date indicates that this alternate method of operation
is free of reverse photoeffect.
It has been well established for the past two or three years that, unless very perfect
vacuum conditions are obtained, gas atoms will adsorb on a clean tungsten filament until
a considerable fraction of a monolayer is produced there. By suddenly increasing the
temperature of this filament after specific time periods of cold standby, one observes
a "burst" of ion current which is a measure of the rate of adsorption of atoms on a clean
surface. A new phenomenon of this type has been observed recently and may be of some
interest. The experimental arrangement in this case involves the presence of a cold
filament which just before flashing can be made very positive with respect to the ion
collector. No other voltages are needed in the tube. If the filament has been standing
by in high vacuum for some days a very measurable positive ion current is observed as
soon as the filament is flashed to a temperature to the order of 1000°K or higher. The
the effect and likeintroduction of a high pressure of mercury vapor does not enhance
-6
wise the presence of nitrogen at pressure of the order of 10-6 mm seems to have no
After a few hours of baking at about 450 0 C this ion flash is extremely great and
actually yields a much greater number of ions to the ion collector than is the case for
the ordinary flash filament method of observing the presence of gas. We assume that
effect.
one of the more likely atoms that might yield this effect could possibly be sodium.
-7-
This
(I.
PHYSICAL ELECTRONICS)
experimental result is described in detail so that other workers in the field prepared to
identify the presence of ions by mass spectrographic methods might be interested in
looking for a similar phenomenon and may be able to identify the mass associated with
the substance.
W. B. Nottingham, L. Sprague
-8-